Automated segmentation of organs, such as kidneys, from magnetic resonance images

Abstract

A method of segmenting an MR organ volume includes performing regional mapping on the MR organ volume using a spatial prior probability map of a location of the organ to create a regionally mapped MR organ volume, and performing boundary refinement on the regionally mapped MR organ volume using a level set framework that employs the spatial prior probability map and a propagated shape constraint to generate a segmented MR organ volume.

What is claimed is: 
     
       1. A method of segmenting an MR organ volume, comprising:
 performing regional mapping on the MR organ volume using a spatial prior probability map of a location of the organ to create a regionally mapped MR organ volume, wherein the regional mapping comprises:
 preprocessing the MR organ volume by application of total variation regularization to the MR organ volume to create a TV regularized MR organ volume; 
 computing magnitudes of image gradients for the TV regularized MR organ volume; and 
 generating a map of candidate organ regions for the MR organ volume by multiplying the magnitudes of image gradients by the spatial prior probability map; and 
 
 performing boundary refinement on the regionally mapped MR organ volume using a level set framework that employs the spatial prior probability map and a propagated shape constraint to generate a segmented MR organ volume. 
 
     
     
       2. The method according to  claim 1 , wherein the MR organ volume is an MR kidney volume, wherein the regionally mapped MR organ volume is a regionally mapped MR kidney volume, and wherein the segmented MR organ volume is a segmented MR kidney volume. 
     
     
       3. The method according to  claim 2 , wherein the spatial prior probability map is a spatial prior probability map of the location of kidneys in a number of abdominal MR images. 
     
     
       4. The method according to  claim 1 , wherein the MR organ volume is an MR kidney volume and wherein the method further includes separating the map of candidate organ regions into a right kidney region and a left kidney region. 
     
     
       5. The method according to  claim 1 , wherein the boundary refinement comprises:
 iteratively determining an evolved contour for the MR organ volume using the map of candidate organ regions and the level set framework; and 
 performing morphological closing on the evolved contour to generate the segmented MR organ volume. 
 
     
     
       6. The method according to  claim 1 , wherein the propagated shape constraint enforces organ contours in neighboring MR images of the MR organ volume. 
     
     
       7. The method according to  claim 1 , further comprising generating and displaying a segmented output image using the segmented MR organ volume. 
     
     
       8. A non-transitory computer readable medium storing one or more programs, including instructions, which when executed by a computer, causes the computer to perform the method of  claim 1 . 
     
     
       9. A computerized system for segmenting an MR organ volume, comprising:
 a processing apparatus, wherein the processing apparatus includes:
 a spatial prior probability map component that includes a spatial prior probability map of a location of the organ; 
 a regional mapping component configured for creating a regionally mapped MR organ volume by performing regional mapping on the MR organ volume using the spatial prior probability map, wherein the regional mapping component is structured and configured to, preprocess the MR organ volume by application of total variation regularization to the MR organ volume to create a TV regularized MR organ volume; compute magnitudes of image gradients for the TV regularized MR organ volume; and generate a map of candidate organ regions for the MR organ volume by multiplying the magnitudes of image gradients by the spatial prior probability map; and 
 a boundary refinement component configured for generating a segmented MR organ volume by performing boundary refinement on the regionally mapped MR organ volume using a level set framework that employs the spatial prior probability map and a propagated shape constraint, wherein the map of candidate organ regions is used by the boundary refinement component. 
 
 
     
     
       10. The system according to  claim 8 , further comprising a display structured to receive the segmented MR organ volume and generate and display an output image based on the segmented MR organ volume. 
     
     
       11. The system according to  claim 9 , wherein the MR organ volume is an MR kidney volume, wherein the regionally mapped MR organ volume is a regionally mapped MR kidney volume, and wherein the segmented MR organ volume is a segmented MR kidney volume. 
     
     
       12. The system according to  claim 11 , wherein the spatial prior probability map is a spatial prior probability map of the location of kidneys in a number of abdominal MR images. 
     
     
       13. The system according to  claim 9 , wherein the MR organ volume is an MR kidney volume and wherein the regional mapping component is structured and configured to separate the map of candidate organ regions into a right kidney region and a left kidney region. 
     
     
       14. The system according to  claim 9 , wherein the boundary refinement component is structured and configured to:
 iteratively determine an evolved contour for the MR organ volume using the map of candidate organ regions and the level set framework; and 
 perform morphological closing on the evolved contour to generate the segmented MR organ volume. 
 
     
     
       15. The system according to  claim 9 , wherein the propagated shape constraint enforces organ contours in neighboring MR images of the MR organ volume.